Virtual Machine Update via Raw Block Storage Detachment

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Updating virtual machines on cloud services with distributed databases is cumbersome and time-consuming, requiring rebuilding of database nodes, which is inefficient and often manual.

Innovation Solution

Implementing systems and methods that allow for the re-initiation of virtual machine instances without rebuilding distributed database nodes by using raw block storages to preserve and transfer data during the update process, enabling faster updates and automation of the process.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If virtual machines are updated by terminating old instances and initiating new instances, then the virtual machines can be updated to the latest application version, but substantial time is required to re-create the distributed database through manual backup and rebuild or duplication across new instances

Engineering Contradiction:
Improveapplication versionVSAvoidupdate time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent separates the distributed database storage from the virtual machine instances by using raw block storage devices. Each database node is stored as an independent block device that can be detached from one virtual machine and attached to another without affecting the database integrity. This segmentation allows the database storage layer to be independent from the compute layer, enabling rapid VM updates without database reconstruction.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary actions by maintaining database nodes in raw block storage format that are pre-configured and ready for attachment. When VM updates are needed, the block storage devices are already prepared and can be immediately attached to new VM instances, eliminating the need for time-consuming database reconstruction processes during the update.

Inventive Principle:
Principle #10Preliminary action

2Adaptability or versatility

If distributed databases are implemented on virtual machines, then cloud services can provide database functionality, but the difficulty of updating virtual machines increases significantly

Engineering Contradiction:
Improvedatabase functionalityVSAvoidupdate complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent extracts the distributed database storage functionality from the virtual machine instances and places it in raw block storage devices managed at the infrastructure layer. This extraction allows database functionality to be maintained while simplifying the VM update process, as the database nodes remain in storage and are simply reattached to new VM instances rather than being rebuilt within the VMs themselves.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent introduces raw block storage devices as an intermediary between the virtual machines and the distributed database nodes. These block storage devices act as a mediator that decouples the database storage from the compute resources, allowing VMs to be updated without affecting database integrity or requiring complex coordination between database reconstruction and VM lifecycle management.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Reliability

If manual backup and rebuild of database nodes is performed during virtual machine updates, then database continuity can be maintained, but the process is time-consuming and labor-intensive

Engineering Contradiction:
Improvedatabase continuityVSAvoidupdate speed
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent uses raw block storage devices that can be copied or cloned rapidly across different virtual machine instances. Instead of performing manual backup and rebuild operations, the database nodes are stored as copyable block devices that can be instantly replicated and attached to new VM instances, maintaining database continuity while dramatically increasing update speed and reducing manual intervention.

Inventive Principle:
Principle #26Copying

Data Source

PatentUS10977069B2Systems and methods for performing virtual machine updates without rebuild of distributed databases thereon
Publication Date: 2021.04.13 CAPITAL ONE SERVICES LLC
  • US10977069B2 patent drawing
  • US10977069B2 patent drawing
  • US10977069B2 patent drawing

AI summary

The present disclosure relates to systems and methods for updating virtual machines without rebuilding distributed databases thereon. In one example, a system for updating a virtual machine on a cloud service may include at least one processor configured to: generate and transmit, to the cloud service, a first command to disconnect a virtual storage from a first virtual machine; after the virtual storage is disconnected, generate and transmit, to the cloud service, a second command to offline the first virtual machine; generate and transmit, to the cloud service, a third command to build a second virtual machine; after the second virtual machine is built, generate and transmit, to the cloud service, a fourth command to connect the virtual storage to the second virtual machine; and generate and transmit, to the cloud service, a fifth command to online the second virtual machine with the address of the virtual machine.